1. Field of the Invention
The present invention relates to a focus detection apparatus that extracts feature portions of an object of shooting, and performs a focus detection based on the extraction result, and a control method therefor.
2. Description of the Related Art
In an automatic focus (AF) control of video cameras, a TV-AF method to generate an AF evaluation value indicating sharpness (contrast state) of a video signal generated by an image sensor, and to search for a position of a focus lens at which the AF evaluation value becomes maximal, has become mainstream. However, there is a problem that when, for example, photographing a person, instead of focusing on the person of a main object, a background comes into focus due to a relationship of a contrast.
To solve such a problem, an image pickup apparatus having face recognition functions is known. For example, an image pickup apparatus (see, for example, Japanese Patent Application Laid-Open No. 2006-227080) sets a focus detection region for detecting an in-focus state in a region including a recognized face region, and performs a focus detection. Alternatively, an image pickup apparatus (see, for example, Japanese Patent Application Laid-Open No. 2001-215403) detects a person's eyes, and performs a focus detection with reference to the detected eyes.
In the focus detection using the face recognition functions described above, when a face has been recognized, a focus adjustment is performed in the recognized face portion, so as to immediately deal with an object person as intended by a user. However, in a case of a moving image pickup process, various objects may often be present on a screen coming one after another. In such a case, if a focus adjustment is performed on a recognized face every time, when a passer-by who happens to pass is temporarily present at an edge of the screen, the focus adjustment will be performed on this person. Eventually, the focus adjustment will not be performed on the person as intended by the user. In other words, generally when picking up a moving image, the user often captures the image so as to arrange the object in the center of the screen. However, when a face is recognized at an edge of the screen, the focus adjustment may be performed on the face at the edge. As a result, the focus adjustment is not performed on the object intended by the user.
Further, in the above-described situation, a face of the object person temporarily present at an edge of the screen is recognized, and the focus adjustment on the face of the person at the edge is going to be performed, but, after that, the person can instantly disappear from the screen. In this case, such an unstable focus adjustment will be performed that the focus adjustment will again be performed on the object located in the center of the screen, which will eventually annoy the user.